HAADF-STEM and DFT investigations of the Zn-containing β″ phase in Al–Mg–Si alloys

Abstract

The Zn-containing β" phase in Al-Mg-Si alloys has been investigated by aberration corrected high angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), combined with density functional theory (DFT) calculations. The mean intensity of one Si site of the β" phase is higher than the other Si sites, suggesting partial Zn occupancy. DFT studies support that this Si site is competitive for Zn incorporation. While HAADF-STEM image simulations show an influence of the Zn distribution along the β" main growth direction, total energy calculations predict a weak Zn-Zn interaction. This suggests that Zn atoms are not clustering, but uniformly distributed along the atomic columns. The Zn incorporation has a weak influence on the β" phase where Zn is admitted as a "defect" according to the DFT studies.

Cu atoms suppress misfit dislocations at the β″/Al interface in Al–Mg–Si alloys

“…Without Cu introductions, the simulated HAADF-STEM intensities for the bulk and {-320}Al interface Si3/Al sites differ only weakly (see Figure 2). For a specimen thickness of 40 nm (which is represented by Figure 1 (b)), the interface intensities are higher by ~12% (due to the misfit strain and/or interface electronic effects [14]). Note that the intensity difference is twice as high in the experiment (~24%).…”

“…[14] and references cited therein. The analysis of the experimental HAADF-STEM intensities was conducted by measurement of integrated column intensities for each column, assuming fixed column atom density.…”

“…Cu binding energies at the β"/Al interface were computed as the energy change upon moving this "defect" from a fcc Al 108 atom supercell to a Si3/Al site in the interface cell (and Al the other way). By comparison, the (bulk β") "replacement energies" [14] typically involve introduction of additional Al defects for the reference configurations, generally comparing the Cu introduction with the best solute atom alternative. Calculations used a 234 eV PW cut-off, with compatible Monkhorst-Pack k-point grids for the various cells ( (12,6,14) for the conventional β" cell).…”

“…By comparison, the (bulk β") "replacement energies" [14] typically involve introduction of additional Al defects for the reference configurations, generally comparing the Cu introduction with the best solute atom alternative. Calculations used a 234 eV PW cut-off, with compatible Monkhorst-Pack k-point grids for the various cells ( (12,6,14) for the conventional β" cell). Without Cu introductions, the simulated HAADF-STEM intensities for the bulk and {-320}Al interface Si3/Al sites differ only weakly (see Figure 2).…”

“…For the simulations, adjustable parameters (slice thickness, convergence semi-angle, HAADF collector angles, frozen phonon configurations and Debye-Waller factors) defined in Ref. [14] were unaltered. The chosen Debye-Waller factor for the Cu site matched the elemental phase value at 300 K: 0.5747 Å 2 [16].…”